单神经元自适应PID控制算法及仿真研究.docVIP

单神经元自适应PID控制算法及仿真研究 摘要 本文研究了神经网络PID算法的最简单结构--单神经元自适应PID算法，利用VB语言编写仿真程序，对常见一阶、二阶对象进行了神经网络PID仿真研究，并分析单神经元自适应PID控制的各参数对控制效果的影响，同时分析了单神经元PID较PID控制的优点。主要研究内容有以下几方面： 概述了本课题研究的背景PID控制和智能控制理论发展状况 传统PID控制的原理的介绍及仿真，主要介绍了PID的控制原理以及参数整定的工程方法，并用MATLAB对其进行仿真。 2、单神经元自适应PID控制的研究。主要介绍了单神经元自适应PID控制的原理和单神经元自适应PID控制算法，使用VB编写程序，对一阶、二阶系统分别用单神经元自适应PID算法和传统PID算法进行仿真，并且分析了单神经元自适应PID控制中的比例、积分、微分学习率对系统性能的影响。 3、将单神经元自适应PID算法进行比较分析关键：神经网络， The Simulation Research of Single Neural Adaptive PID Control Algorithm ABSTRACT In this paper, the single neuron adaptive PID algorithm was studied. Based on VB, the one and two order object was simulated with the neural PID algorithm. The effect of the single neuron adaptive PID control parameters on the control performance was analyzed. And its performance was compared with PID controllers. The main research contents are as followings: 1. The background of this topic, development of PID controller and intelligent control theory was summarized. 2. The traditional PID control principle and the engineering method of PID parameter tuning was introduced .Simulation research was done using MATLAB. 3. The single neuron adaptive PID control principle and single neuron adaptive PID control algorithm were researched. The simulation program was written using VB programming language .The single neuron PID algorithm was used to control the one order, second order process with time-delay. The control performance of single neuron PID with traditional PID was compared. The proportional, integral, differential learning rate on system performance was analyzed. 4. The single neuron adaptive PID control algorithm is superior to the traditional PID control algorithm in overall by comparing and analyzing between the single neuron adaptive PID algorithm with the traditional PID control algorithm. It is helpful for control system to improve the control quality. It has strong robustness, and the environment have limited influence on the system. It is a very promising cont